Below are interactive 3D plots of the foot and mouth disease data.

Code

The code to produce the above interactive is:

library(sparr)
library(rgl)
library(misc3d)

# Load data and extract cases and controls
data(fmd)
fmd_case <- fmd$cases
fmd_cont <- fmd$controls

# fit case and control densities, and estimate risk
f_breve <- spattemp.density(fmd_case, h = 3 , lambda = 9 , tlim = c(10, 230), sres = 32, tres = 32)
g_tilde <- bivariate.density(fmd_cont, h0 = 3, res = 32)
rho_breve <- spattemp.risk(f = f_breve, g = g_tilde, tolerate = TRUE)

# extract evaluation grid
tres <- length(rho_breve$rr)
x <- rho_breve$rr[[1]]$xcol
y <- rho_breve$rr[[1]]$yrow
z <- as.numeric(names(rho_breve$rr))
w <- vertices(Window(fmd_case))

# get log-risk surface as 3D array, and replace NAs
rr <- array(NA, dim = c(ncol(rho_breve$rr[[1]]), nrow(rho_breve$rr[[1]]), tres))
for(i in 1:tres)
  rr[,,i] <- t(as.matrix(rho_breve$rr[[i]]))
rr[is.na(rr)] <- min(rr, na.rm = TRUE) - 1

# get P-value surface as 3D array, replacing NAs
pp <- array(NA, dim = c(ncol(rho_breve$P[[1]]), nrow(rho_breve$P[[1]]), tres))
for(i in 1:tres)
  pp[,,i] <- t(as.matrix(rho_breve$P[[i]]))
pp[is.na(pp)] <- max(pp, na.rm = TRUE) + 1

# setup a 2x1 graphic
mfrow3d(nr = 1, nc = 2, sharedMouse = FALSE)

# colours to use
cols <- spatstat.options("image.colfun")(3) 

# plot left panel
plot3d(fmd_case$x, fmd_case$y, marks(fmd_case), type = "n",
       xlim = range(x), ylim = range(y), zlim = range(z), box = FALSE,
       axes = FALSE, xlab = "Easting", ylab = "Northing", zlab = "")
lines3d(c(w$x, w$x[1]), c(w$y, w$y[1]), rep(z[1], length(w$x) + 1))
lines3d(c(w$x, w$x[1]), c(w$y, w$y[1]), rep(max(z), length(w$x) + 1))
contour3d(x = x, y = y, z = z, f = rr, level = c(0, 1, 2),
          add = TRUE, alpha = c(0.1, 0.4, 0.7), color = cols)

# plot right panel
plot3d(fmd_case$x, fmd_case$y, marks(fmd_case), type = "n",
       xlim = range(x), ylim = range(y), zlim = range(z), box = FALSE,
       axes = FALSE, xlab = "Easting", ylab = "Northing", zlab = "")
lines3d(c(w$x, w$x[1]), c(w$y, w$y[1]), rep(z[1], length(w$x) + 1))
lines3d(c(w$x, w$x[1]), c(w$y, w$y[1]), rep(max(z), length(w$x) + 1))
contour3d(x = x, y = y, z = z, f = pp, level = c(0.05, 0.0001),
          add = TRUE, alpha = c(0.2, 0.5), color=c("yellow", "red"))
points3d(fmd_case$x, fmd_case$y, marks(fmd_case))

# save widget
rglwidget(width = 960, height = 540)